I wasn’t surprised to learn that the number of confirmed exoplanets had finally topped 6,000, a fact recently announced by NASA. After all, new worlds keep being added to NASA’s Exoplanet Science Institute at Caltech on a steady basis, all of them fodder for a site like this. But I have to admit to being startled by the fact that fully 8000 candidate planets are in queue. Remember that it usually takes a second detection method finding the candidate world for it to move into the confirmed ranks. That 8000 figure shows how much the velocity of discovery continues to increase.
The common theme behind much of the research is often cited as the need to find out if we are alone in the universe. Thus NASA’s Dawn Gelino, head of the agency’s Exoplanet Exploration Program (ExEP) at JPL:
“Each of the different types of planets we discover gives us information about the conditions under which planets can form and, ultimately, how common planets like Earth might be, and where we should be looking for them. If we want to find out if we’re alone in the universe, all of this knowledge is essential.”
I sometimes think, though, that the emphasis on an Earth 2.0 is over-stated. The search for other life is fascinating, but deepening our scientific knowledge of the cosmos is worthwhile even if we learn we are alone in the galaxy. What nature creates in bewildering variety merits our curiosity and deep study even on barren worlds. With ESA’s Gaia and NASA’s Roman Space Telescope in the mix, exoplanet detections will escalate dramatically. And a little further down the road is the Habitable Worlds Observatory, assuming we have the good sense to green-light the project and build it.
Bear in mind that almost all the known exoplanets are within a few thousand light years of Sol. We are truly awash in immensity. If there will ever be a complete catalog of the Milky Way’s planets, it will likely be from a Kardashev Type III civilization immensely older than ourselves. It’s fascinating to think that such a catalog might already exist somewhere. But it’s also fascinating to consider that we may be alone, which raises all kinds of questions about abiogenesis and the possible lifetime of civilizations.
Image: Scientists have found thousands of exoplanets (planets outside our solar system) throughout the galaxy. Most can be studied only indirectly, but scientists know they vary widely, as depicted in this artist’s concept, from small, rocky worlds and gas giants to water-rich planets and those as hot as stars. Credit: NASA’s Goddard Space Flight Center.
The idea of a large and growing catalog of exoplanets is the kind of thing I used to dream about as a kid reading science fiction magazines. Now we’re on the cusp of biosignature detection capabilities via the deep study of exoplanet atmospheres. Fewer than a hundred exoplanets have been directly imaged, a number that is likewise expected to rise with the help of the Roman instrument’s coronagraph. With new tools to better block out the overwhelming glare of the host star, we’ll be seeing gas giants in Jupiter-like orbits. That in itself is interesting – how many exoplanet systems have gas giants in such positions? Thus far, the Solar System pattern is rarely replicated.
The Fortunes of K2-18b
The sheer variety of planetary systems brings even more zest to this work. Consider the planet K2-18b, so recently in the news as the home of a possible global ocean, and one with prospects for life given all the parameters studied by a team at the University of Cambridge. It’s a fabulous scenario, but now we have a new study that questions whether sub-Neptunes like this are actually dominated by water. Caroline Dorn (ETH Zurich), co-author of the paper appearing in The Astrophysical Journal, believes that water on sub-Neptunes is far more limited than we have been thinking.
Here’s another demonstration of how our Solar System is so unlike what we’re finding elsewhere. Lacking a sub-Neptune among our own planets, we’re learning now that such worlds – larger than Earth but smaller than Neptune and cloaked in a thick atmosphere abundant in hydrogen and helium – are relatively common in our galaxy as, for that matter, are higher density but smaller ‘super-Earths.’ A global ocean seems to make sense if a sub-Neptune formed well beyond the snowline and brought a robust inventory of ice with it as it migrated into the warmer inner system. Indeed, the term Hycean (pronounced HY-shun) has been proposed to label a sub-Neptune planet with a deep ocean under an atmosphere rich in hydrogen.
The new paper examines the chemical coupling between the planet’s atmosphere and interior, with the authors assuming an early stage of formation in which sub-Neptunes go through a period dominated by a magma ocean. The hydrogen atmosphere helps to maintain this phase for millions of years. And the problem is that magma oceans have implications for the water content available. Using computer simulations to model silicates and metals in the magma, the team studied the chemical interactions that ensue. Most H20 water molecules are destroyed, with hydrogen and oxygen bonding into metallic compounds and disappearing deep into the planet’s interior.
From the paper:
Our results, which focus on the initial (birth) population of sub-Neptunes with magma oceans, suggest that their water mass fractions are not primarily set by the accretion of icy pebbles during formation but by chemical equilibration between the primordial atmosphere and the molten interior. None of the planets in our model, regardless of their initial H2O content, retain more than 1.5 wt% water after chemical equilibration. This excludes the high water mass fractions (10–90 wt%) invoked by Hycean-world scenarios (N. Madhusudhan et al. 2021), even for planets that initially accreted up to 30% H2O by mass. These findings are consistent with recent studies suggesting that only a small amount of water can be produced or retained endogenously in sub-Neptunes and super-Earths.
The work analyzes 19 chemical reactions and 26 components across the range of metals, silicates and gases, with the core composed of both metal and silicate phases. A computer model known as New Generation Planetary Population Synthesis (NGPPS) combines planetary formation and evolution and meshes with code developed for global thermodynamics. Thus a population of sub-Neptunes with magma oceans is generated, and consistently primordial water is destroyed by chemical interactions.
Image: This is Figure 3 from the paper. Caption: Envelope H2O mass fraction as a function of semimajor axis… The left panel shows planets that predominantly formed inside the water ice line; the right panel shows those that formed outside. Classification is based on the accreted H2O mass fraction, with a threshold set at 5% of the total planetary mass. The colorbar indicates the molar bulk C/O ratio. Planets formed inside the ice line are systematically depleted in carbon due to the lack of volatile ice accretion and exhibit higher envelope H2O mass fractions. In contrast, planets formed beyond the ice line retain lower H2O content despite higher bulk volatile abundances. Each pie chart shows the mean mass fraction of hydrogen in H2 (gas), H (metal), H2 (silicate), H2O (gas), and H2O (silicate), normalized to the total mean hydrogen inventory for each population. Only components contributing more than 5% are labeled. Planets that formed beyond the ice line store most hydrogen as H2 (gas) and H (metal), while those that formed inside the ice line retain a larger share of hydrogen in H (metal), H2 (silicate), and H2O (gas + silicate).. Credit: Werlen et al.
This is pretty stark reading if you’re fascinated with deep ocean scenarios. Here’s Dorn’s assessment:
“In the current study, we analysed how much water there is in total on these sub-Neptunes. According to the calculations, there are no distant worlds with massive layers of water where water makes up around 50 percent of the planet’s mass, as was previously thought. Hycean worlds with 10-90 percent water are therefore very unlikely.”
The paper is suggesting that we re-think what had seemed an obvious connection between planet formation beyond the snowline and water in the atmosphere. Instead, the interplay of magma ocean and atmosphere may deliver the verdict on the makeup of a planet. By this modeling, planets with atmospheres rich in water are more likely to have formed within the snowline. That leaves rocky worlds like Earth in the mix, but raises serious doubts about the viability of water in the sub-Neptune environment.
From the paper:
Counterintuitively, the planets with the most water-rich atmospheres are not those that accreted the most ice, but those that are depleted in hydrogen and carbon. These planets typically form inside the ice line and accrete less volatile-rich material. While some retain significant atmospheric H2O, the high-temperature miscibility of water and hydrogen likely prevents the presence of surface liquid water—even on these comparatively water-rich worlds.
This has broad implications for theories of planet formation and volatile evolution, as well as for interpreting exoplanet atmospheres in the era of the James Webb Space Telescope (JWST), the Extremely Large Telescope (ELT), ARIEL, the Habitable World Observatory (HWO), and the Large Interferometer for Exoplanets (LIFE). It also informs atmospheric composition priors in interior characterization of transiting planets observed by Kepler, TESS, CHEOPS, and PLATO with radial velocity (RV) or transit timing variation (TTV) constraints.
K2-18b will obviously receive continued deep study. But with an aggregate 14,000 exoplanets confirmed or listed as candidates, consider how overwhelmed our instruments are by sheer numbers. To do a deep dive into any one world demands the shrewdest calculations to find which exoplanets are most likely to reward the telescope time. This aspect of target selection will only get more critical as we proceed.
The paper is Werlen et al., “Sub-Neptunes Are Drier than They Seem: Rethinking the Origins of Water-rich Worlds,” Tje Astrophysical Journal Letters Vol. 991, No. 1 (18 September 2025), L16 (full text).
Looks like a job for AI.
IIUC, hycean worlds probably don’t exist if the model is correct. This reminds me of the belief in a tropical, wet Venus rather like the tropical, wet Venus that gave way to a hot, dry Hades instead.
However, this only applies to sub-Neptunes, and not the smaller rocky worlds that may have a variable amount of primordial and cometary water, from being dry, like the fictional Arrakis, to water worlds.
While the sub-Neptunes may be relatively dry, could the larger moons of ice and gas giants that have migrated into the HZ have deep oceans, if their mass were great enough to retain their water and an atmosphere as the moon warmed up in the HZ? [Such a moon would have to be larger and more massive than Ganymede, the largest moon in our system.]
Good point Alex Tolley. There is more that has to be considered like Jeans escape which is why planets inside the ice line in our solar system have a lot of hydrogen and helium since the temperature is lower and an earth sized planet or a planet with more mass can retain much more atmosphere than in the inner solar system which is ignored in this study. The mass is what counts and not the chemistry, so there are not any gray areas or lope hole to fit through. I corroborated this with Open AI Chat GPT.
Consequently, it does not invalidate in any way the super Earth ocean world or the Sub Neptune or Hycean world which has H and He over a pressurized ocean. It’s not the chemistry, but the mass and the mass of the atmosphere. A mini Neptune would still have H, He on top of pressurized, frozen water because of the pressure it becomes like ice even if it outside the ice line near the star. We get the same thing with Uranus and Neptune with pressure ionized ice layer under the gas, the H2 and He. Ibid..
This of course goes back to the fundamentals of Geology, The Goldschmidt classifications the lithosphere and lithophiles, Chalcophiles, atmophiles stay near the surface and the Siderophiles and heavy metals sink to the core called differentiation.
How Interstellar Objects Similar to 3I/ATLAS Could Jump-Start Planet Formation Around Infant Stars.
Susanne Pfalzner
Jülich Supercomputing Centre, Forschunsgzentrum Jülich, Jülich, Germany
Interstellar objects like 3I/ATLAS that have been captured in planet-forming discs around young stars could become the seeds of giant planets, bypassing a hurdle that theoretical models have previously been unable to explain.
Interstellar objects are asteroid- and comet-like bodies that have been ejected from their home system and now wander through interstellar space, occasionally encountering other star systems. Since 2017 astronomers have detected three interstellar objects passing through our Solar System: 1I/’Oumuamua, 2I/Borisov and most recently 3I/ATLAS, discovered in summer 2025.
However, interstellar objects may be more influential than they at first appear to be, says Professor Susanne Pfalzner of Forschungszentrum Jülich in Germany, who presents her new findings on the subject at this week’s EPSC-DPS2025 Joint Meeting in Helsinki.
“Interstellar objects may be able to jump start planet formation, in particular around higher-mass stars,” said Pfalzner.
Planets form in dusty discs around young stars through a process of accretion, which according to theory involves smaller particles come together to form slightly larger objects, and so on until planet-sized bodies have assembled. However, theorists struggle to explain how anything larger than a metre forms through accretion in the hurly-burly of a planet-forming disc around a young star – in computer simulations, boulders either bounce off each other or shatter when they collide rather than sticking together.
Interstellar objects can potentially bypass this problem. Pfalzner’s models show how the dusty planet-forming disc around each young star could gravitationally capture millions of interstellar objects the size of 1I/’Oumuamua, which was estimated to be around 100 metres long.
“Interstellar space would deliver ready-made seeds for the formation of the next generation of planets,” said Pfalzner.
If interstellar objects can act as the seeds of planets, it also solves another mystery. Gas giant planets like Jupiter are rare around the smallest, coolest stars, which astronomers refer to as ‘M dwarfs’. They are more commonly found around more massive stars similar to the Sun. The problem, though, is that planet-forming discs around Sun-like stars have a lifetime of about two million years before dissipating and it’s very challenging to form to form gas giant planets on such a short timescale. However, if captured interstellar objects are present as seeds onto which more material can accrete, it speeds the process of planet formation up and giant planets can form in the lifetime of the disc.
“Higher-mass stars are more efficient in capturing interstellar objects in their discs,” said Pfalzner. “Therefore, interstellar object-seeded planet formation should be more efficient around these stars, providing a fast way to form giant planets. And, their fast formation is exactly what we have observed.”
Pfalzner says that her next steps are to model the success rate of these captured interstellar objects – investigating how many of the millions of captured interstellar objects are able to form planetary bodies, and whether they are captured evenly across a planet-forming disc, or whether they are concentrated in certain areas that could become hotspots for planet-birth.
Simulations of the capture of ISOs in molecular clouds indicate that ISOs are predominately captured by massive stars or in star cluster environments (Pfalzner et al. 2021).
Exoplanet statistics show considerable differences between the planetary systems around high-mass and those around low-mass stars. Main-sequence FGK stars host more larger planets than low-mass stars, whereas M dwarfs host about a factor of three more small planets
(Mulders et al. 2021).
However, the mass is not simply redistributed into more smaller planets. Surprisingly, the average heavy-element mass decreases with increasing stellar mass. Thus, despite M star discs containing ten times less mass, they are nearly 20 times as efficient as F stars in converting the disk’s heavy-element content into planetary material. Besides, high-mass stars have, on average, much shorter disc lifetimes (1-3 Myr) than low-mass stars (5-10 Myr). Thus, giant planets have to form on extremely short time scales.
Could this be the real reason for hot Jupiters…
“Simulations of the capture of ISOs in molecular clouds indicate that ISOs are predominantly captured by massive stars or in star cluster environments.”
If formed massive stars solar systems pass through molecular clouds or in star cluster environments, could ISOs replenish dry super Earths and Sub Neptunes and create Hycean worlds.
Manna from heaven.
The problem here is the velocity of these objects, they would simply move through the newly forming systems and out the other side.
Michael, you may want to look at this ladies research. She may know a lot more then you do.
https://scholar.google.com/citations?user=dplYCPwAAAAJ&hl=en
Assorted comments.
I have been pronouncing hycean wrong.
Using probes, a SFI could catalog all planets without reaching K1 status. Certainly much faster that it could colonize a galaxy.
To put the 1.5% water by total mass threshold for sub-Neptunes into perspective, Earth’s is 0.02 – 0.023%.
The extremely deep oceans previously modeled made abiogenesis unlikely. If true, wouldn’t this modeling increase the probability that sub-Neptunes are habitable?
Same here. I have been pronouncing it as “Hy-see-an”.
Wikipedia entry for Hycean pronunciation:
HY-shən
Using their pronunciation key:
ə about, comma /ə/
It seems it depends on how one pronounces about, comma. It can be “ay” (aybout), “a” (hat), or even “uh”, or anything in between ;-)
I don’t know if the examples are from which parts of the English-speaking world they are from, or which accents. The respelling approach works better for me, as UK English puts very different places for the stresses, which makes many US place names difficult to pronounce until heard from a native. [It also makes a mess of British place names with French (Norman) origins, e.g., ““Beaulieu” is pronounced “Bewley” in England, which makes anyone with even a smattering of French raise an eyebrow. Conversely, most North Americans cannot correctly pronounce UK placenames e.g., “Edinburgh” or “Greenwich”. ]
I’ll say a prayer to Lexosaurus, the god of dictionaries, that it compel humans to add an etymological link for every mention of a neologism for the first decade of its existence. That way we’d understand why hycean is pronounced the way it is.
The inhabitants should have the last say. But I subvocalized it several times when I read about these matters, so I hope that I hadn’t offended.
It rhymes with ocean.
@TPS
Note the subtle differences between just the UK and US pronunciations of “ocean“.
I have heard it pronounced as “Oh-she-an” on TV before, but admittedly, not “Oh-see-an”.
Quote by authors of this paper: “Our results, which focus on the initial (birth) population of sub-Neptunes with magma oceans, suggest that their water mass fractions are not primarily set by the accretion of icy pebbles during formation but by chemical equilibration between the primordial atmosphere and the molten interior.” This is where the real problem of this paper begins is since differentiation always must come last since the accretion disk always must come first long with the building of the planets through collisions of protoplanets and then comes the differentiation of the elements during the molten phase. Even though Earth collided with Thea to make a molten surface, the water amounts were already preset and therefore regained it’s oceans.
Consequently, there is a lot of physics in planetology which is complicated and draws upon many fields and subjects in geology, chemistry, astrphysics, meteorology, etc and we can’t always force everything into a simple idea like the snow line, etc. We can’t change the system or rewrite any of those first principles with falling into error or correctly matching observations.
With Venus we know that it had to have an ocean before it moved out of the life belt when the Sun increases in brightness seven percent every billion years. The evidence is based on the deuterium water ratio in Venus atmosphere which is hundreds to thousands percent higher than Earth. The heaver DH20 gets left behind when the light H20 is stripped into hydrogen and oxygen and the heavy hydrogen gets left behind as HDO and D2O. Open AI Chat GPT Much oxygen is also lost because there should be a false positive of much oxygen as the result of its splitting apart with UV, but very little oxygen remains and we knew this with looking at Venus atmosphere through a spectrometer through a telescope at Venus atmosphere before any space probes were sent there which showed no water vapor and no oxygen. Furthermore, when the oceans evaporated into space and not any plate tectonics or plate movement there could not be any carbon cycle and oceans to help remove the carbon dioxide, it kept increasing due to volcanism and we have an atmosphere mostly of carbon dioxide.
@Geoffrey
Would we also expect the O_18/O_16 ratio to also be much higher for the oxygen in the CO2, and the rocks, e.g. SiO2 as the O_16 would have been more likely to escape, or was all the oxygen retained?
This paper sheds some light on this, although I confess to not understanding it well on a skim: https://planetary.aeronomie.be/ProjectDir/Publications/Iwagami_15.pdf
The oxygen isotope ratios tell where a planet came from in the solar system. Earth Moon oxygen isotope ratios are nearly identical so moon formed from a giant impact with Earth and Theia. The can’t be used Also O16/O18 ratios tell the temperature, rainfall and ice covering. These support the idea of the Milankovitch cycles and it’s ice and and warm period cycles.
It is the D/H ratios that tell us the amount of water that has been lost atmospheric escape through solar wind, etc.
To be more specific the water was delivered to Earth by the collisions of the planetesimals which still came from the protoplanetary accretion disk through its condensation of the gas, dust and pebbles. The oxygen ratios are different for comets are 2 to 3x higher than Earth which is evidence that Earth did not get its water from comets and outer solar system where the comets came, the Oort Cloud comets and many Kuiper Belt comets). Also the D/H ratio in Carbonaceous chondries have similar D/H ratio to Earth, so they formed near the Earth. Comets have a higher Nitrogen ratio than Earths. There is also the noble gases Xe, Kr, Ar isotpes are isotopicially fractionated indication a complex delivery the planetesimals. Open AI Chat GPT.
I have read that an isotopic analysis indicates that about 1/2 the water on Earth is primordial and 1/2 from impactors, likely comets. Is that incorrect? I cannot recall if the water includes water below the crust or not.
Quote by Alex Tolley: “I have read that an isotopic analysis indicates that about 1/2 the water on Earth is primordial and 1/2 from impactors, likely comets. Is that incorrect?” No. Comets follow long cycles so the probability of them hitting the Earth is much lower than nearby carbonaceous Chondrites and the nearby accretion disk from the solar nebula which formed the protoplanetesimals. The low D/H ratio of the Chondrites matches the Earths water. Astrophysicists have not completely ruled out comets and argue a minor contribution to Earth’s water. Open AI Chat GPT.
Quote by Chat GPT: “Comets, especially the Oort Cloud ones, generally have much higher D/H ratios than Earth water. A few Jupiter-family comets measured (e.g. 103P/Hartley 2, 46P/Wirtanen) have ratios close to Earth’s, but they’re the minority.”
The stripping is from the solar wind which carries the light isotopes of hydrogen and oxygen more easily than the heaver ones which get left behind in the atmosphere.
I’ve given the paper a quick going over and I’m a little dubious about some of its conclusions. It is, I think, just one factor among many:
First I notice that while there is a consideration of radiative cooling once the planet has formed, I didn’t see a mention of convective cooling, which is far more efficient.
K2-18b has an estimated density of 2.67, which points to it having a deep atmosphere that is a significant proportion of the planet’s mass.
Previous papers on K2-18b calculate from its spectra that it must have a one to one ratio of steam to Hydrogen. This makes a low mass of steam/supercritical water hard to square with need for a deep atmosphere.
Smaller planets of more Earth-like mass (K2-18b is 8.6 x Earth’s mass) may have lower mantle temperatures. The mantle of a planet under a deep steam atmosphere of say 45 kilobars, which is Earth’s mantle at 100 miles deep, and temperature of less than 1000°C will be solid, which would severely limit mixing between the atmosphere and mantle and hence loss of water to the mantle.
For comparison, water makes up 0.023% of Earth’s mass, so even the upper bound of 1.5% for sub-Neptunes would still mean a lot more water than Earth.
Yes I was just thinking that, 1.5% is heaps. You don’t want half your planet’s mass to be water anyway.